Biological and serological properties of staphylococcal enterotoxins exhibit remarkable thermal stability. The slope of the thermal destruction curve (Z value) for enterotoxin A or D is 27-28.degree. C and it is 32.degree. C for enterotoxin B regardless of the initial concentration of enterotoxin, the heating medium, or the assay system used to detect the enterotoxin. Thermal stability of enterotoxins is influenced by the nature of the food, pH, presence of NaCl, etc., and the type of enterotoxin; enterotoxin A is relatively more stable at pH .gtoreq. 6.0 than at pH 4.5-5.5. Enterotoxin D is relatively more stable at pH 4.5-5.5 than at pH .gtoreq. 6.0. Heat treatments commonly used in food processing, i.e., pasteurization (72.degree. C for 25 s) or ultrahigh temperature heating (290.degree. F for 9 s) in a Spiratherm Unit of Fluid whole milk, smoking and heating of cured sausage to 70-100.degree. C, heating of cheddar cheese to 70-90.degree. C are not effective for complete destruction (to levels below detection by the microslide serological assay) of enterotoxin A or D when present initially at levels normally expected to be found in foods (0.5-1.0 .mu.g/100 ml or g). Enterotoxin A or B (30 .mu.g/ml) heated at 100.degree. C for 25 min in saline, containing 2% gelatin and 0.3% proteose peptone and at pH 7.0, retained biological activity toward humans with a suggestion of increased biological activity after heat treatment, especially with enterotoxin A. The time required at a given temperature for complete destruction of enterotoxin in a given food system depends upon the sensitivity of the detection system used and the amount of food sample used for recovery of the enterotoxin. Treatment of heated enterotoxin A or D with urea and its removal by dialysis before serological assay, increased the recovery of enterotoxin by almost 4-fold as compared to controls which were simply heated and were not treated with urea. Enterotoxin A or D present initially at 1 .mu.g/ml in whole milk (using 2 ml for recovery of enterotoxin with urea treatment following heating and assay by the micro slide method of Casman and Bennett) withstood heating of milk at 121.1.degree. C for 15 min. In the absence of urea treatment of heated toxins in milk, both toxins were inactivated (to < 0.02 .mu.g/ml (within 15 min at 121.1.degree. C; between 13.8-14.8 min for toxin A and 12.8-13.8 min for toxin D.